This Small Business Innovative Research Phase I project proposes the development of a computational model, called GeneRx, to incorporate pharmacogenetics and nonlinear adaptive algorithms toward optimizing anti-schizophrenic therapy on a patient specific basis. Preliminary studies on the anti-schizophrenic drug olanzapine show a 40% patient-by-patient error between predicted starting dose and optimal therapeutic dose, using a prototype trained only with patient chart information. This is a significant reduction from the range of starting doses for olanzapine currently used, which is from 1 to 80 mgs/day. Anti-schizophrenic drugs likewise have a large window of therapeutic options, including significant variation in dosages, medications, and combinations of therapies used. Using patient-specific genetic information in conjunction with patient medical chart information, obtained from schizophrenia studies performed by outside sources, we propose to develop computational models to predict efficacy of treatment (i.e. Response/Non-response) to the neuroleptic risperidone. Genetic data for each patient will be acquired by genotyping DNA from the blood samples, scored as single nucleotide polymorphisms (SNPs) present or absent in key schizophrenia-related genes. GeneRx will take a patient's individual genetic, demographic, and environmental variables and predict if the patient will respond initial risperidone drug therapy. Response will be measured by change in standard severity test scores. A more efficient method to prescribe effective anti-schizophrenic pharmaceuticals would expedite recovery, minimize side effects, and reduce medical costs.